Semiconductor integrated circuits contain large numbers of electronic components such as capacitors, diodes, resistors and transistors built on a single chip. Due to the microscopic scale of these circuits, they are susceptible to component defects caused by material impurities and fabrication hazards.
In order to circumvent this problem, chips are built with redundant components and/or circuits that can be switched-in in lieu of corresponding circuits found defective during testing. Usually the switching-out of a defective component or circuit and the switching-in of a corresponding redundant component or circuit is accomplished by using program logic circuits which are activated by altering or programming certain state devices such as fuses or anti-fuses built into the chip circuitry. The process of altering or programing a subset of the available fuses or anti-fuses in order to switch-in the redundant circuitry is also known collectively as programming or repairing.
FIG. 1 illustrates a typical prior art structure for programming a random access memory (RAM) type integrated circuit (IC). The structure involves a bank 1 of state device modules which control the programming logic circuitry 2, which actually makes the switch between defective and redundant circuit portions in the RAM array 3 based on the output of the state device memory bank 1. The bank contains a number of state device circuit modules. Each module contains at least one state device.
State devices are conductivity alterable component. They are manufactured in an initial state, either open (very low conductivity) or closed (high conductivity), and can be altered or programmed to assume the opposite state. Normally, this alteration is only one way, i.e. an altered device cannot be returned to its previous state. Fuses are manufactured closed and are blown open by applying a sufficient current through the device such that resistive heating causes the normally conductive fuse element to melt or explode, thereby forming an open circuit. Anti-fuses such as dielectric capacitors are manufactured open and are blown or programmed to the closed state by applying a sufficient voltage above their breakdown voltage across their terminals. At this voltage the dielectric layer separating the conductive plates of the capacitor ceases to be electrically insulating. The capacitor then permanently forms a closed circuit between its two terminals.
Although many different state device module circuit designs are possible, each must perform two basic functions: First, altering the state device without subjecting the programming logic circuitry to the programming voltage or current, and second, communicating the state of the fuse or anti-fuse to the programming logic circuitry. FIG. 2 shows a simplified state device module circuit containing one fuse-type state device, a fuse 4 which is connected to ground 5 and a node 6. The node is connected through a resistor 7 to output terminal V.sub.out 8. The node is also connected to terminal V.sub.in 9 and a conductive pad 10.
During programming, a high voltage V.sub.bd is applied to the pad using a probe. The current produced by this voltage causes the fuse to blow, creating an open circuit. V.sub.out is protected from V.sub.bd by resistor 7. After programming is complete, V.sub.bd is removed from the pad.
During operation, V.sub.cc is applied to the V.sub.in terminal. If the fuse is open, V.sub.out will show a positive voltage as seen through the resistor. If the fuse is closed, V.sub.out will show ground.
Other circuitry may be provided to limit the current through the fuse during operation. In the circuit of FIG. 2, when the fuse is not blown, a direct path exists from V.sub.in to ground. In the simplest case, a resistor or some other current limiting device placed between the node 6 and V.sub.in would provide this function.
Typically, the voltage required to alter a fuse or other state device is high enough to damage the transistors and capacitors which make up much of the programming logic circuitry and memory arrays on a typical integrated circuit microchip. The high voltage cannot be multiplexed through other circuitry since it would damage the circuitry typically used for this purpose. In order to protect this circuitry, each state device is given a dedicated pad as a point where the high voltage can be exclusively applied, thus avoiding application of this voltage to the sensitive circuits. Since the pads must be large enough to be engaged by a probe, they take up a substantial portion of the available space on each microchip. This limits the number of pads and state devices. In turn, chip designers have devised elaborate schemes and structures to offer the maximum redundancy with the minimum number of state devices.
In addition, it is difficult to develop redundancy structures whereby a single state device in a bank of state devices may be selectively altered using logic circuitry. The problem here is that the transistors and other devices used to make up the logic circuitry are themselves susceptible to failure when directing voltages and currents high enough to alter the state device.
Because of the use of dedicated altering pads, the current prior art solution requires programming prior to packaging. Since the programming pads are covered with passivation prior to assembly, any switch to redundant circuitry must occur before burn-in and thus, delivery to the customer. Currently, failures detected during and after burn-in result in the entire chip being scrapped.
It would be desirable, therefore, to have a state device module which is capable of receiving a low voltage programming signal which allows high voltage to alter the state device without that high voltage being carried through to other circuitry on the chip, both during and after the programming voltage is applied. It is also desirable to be able to perform this programming before and/or after packaging the integrated circuit.